How to use this checklist
This expands the six-step checklist in our AI-shopper-ready guide. Each item follows the same pattern: what it is, why an agent cares, how to check it and who owns it. Work through Part 1 first. Part 2 covers fit information, which few fashion product pages state clearly.
Part 1: Core product data (your job)
These two checks apply to any e-commerce store, fashion or otherwise. They are your theme, your product data and your ongoing catalogue hygiene.
1. Structured core product data
What it is: price, availability, images and product attributes exposed as Schema.org Product/Offer JSON-LD, not only as page copy.
Why it matters: an agent parsing your PDP needs structured fields to reason with. It can infer meaning from prose, but it cannot reliably match "beautiful floaty dress" against a shopper's stated requirements the way it can match a material or offers.price field.
How to verify: run the product page through Google's Rich Results Test and confirm a Product block with a populated offers field, not just warnings.
Owned by: you and your theme.
2. A machine-readable size guide
What it is: your size chart published as an on-page HTML table, measurements in centimetres per size, not an image and not a linked PDF.
Why it matters: a crawler or agent reads text, not pixels. A size guide baked into a graphic is invisible to it no matter how clearly a human can read it.
How to verify: view page source (or use your browser's "select all" on the size guide) and confirm the numbers are selectable text, not part of an <img>.
Owned by: you and your PDP content. Size recommendations start from this chart.
Part 2: Fit information and integrations
Fit is not a field on most product pages. It is a paragraph in a size-chart tab, if it exists at all. The first three checks here are about stating it clearly. The last two are about what to confirm before you rely on an integration, and how to check the guidance against your own returns.
3. Fit notes as plain text per product
What it is: a short, honest statement of how each product fits: true to size, runs small, relaxed cut, stretch, plus the height and size of the model in the photos.
Why it matters: "Runs small" buried in a customer review is hard for a person or an agent to find and act on. The same fact as a line of text on the page is easy to find. Rendered Fits uses your size chart and fit notes to recommend a size, so the recommendation is only as good as these inputs.
How to verify: read the page source and confirm the fit note is text near the size selector or size guide, not part of an image.
Owned by: you. Rendered Fits reads what you provide.
4. Confirm any agent or API integration per store
What it is: a direct way for an assistant or platform to ask your store about fit, if you want one.
Why it matters: availability differs from store to store. An agent that assumes an endpoint exists will get nothing back, and a page that claims one exists without evidence is misleading.
How to verify: discuss the specific store and workflow with the vendor, and get availability, authentication, scope and data handling confirmed in writing before you build against it or describe it publicly. Our own product facts page says the same: it is not a public API contract.
Decided per store. Nothing is assumed to be live.
5. Fit information in Shopify fields you control
What it is: fit attributes (fit type, model height, size worn) held in Shopify metafields you own, under Settings > Custom data > Products.
Why it matters: apps and themes can read a metafield the same way they read other product fields, so the fit note can be shown on the page and reused elsewhere without retyping.
How to verify: in Shopify admin, open Settings > Custom data > Products and check the definitions, then confirm a few products have values.
Owned by: you, in your own store.
6. Check before relying on MCP or similar interfaces
What it is: Model Context Protocol (MCP) and similar interfaces let an assistant call a vendor's tools directly.
Why it matters: they are worth knowing about, but whether any vendor offers one for your store is a per-store question. Do not list an MCP server on your site, or tell customers about one, until you have tested it.
How to verify: ask for the specification, the authentication method and the data returned, then make a test call yourself.
Decided per store. Confirm availability with the vendor first.
7. Size guidance reviewed against your own returns
What it is: a regular look at returns tagged too big or too small, by product.
Why it matters: a size chart nobody has checked against what customers report drifts out of date. Your own return reasons are the best evidence of where a chart or fit note is wrong. Results vary by store and must be measured per store.
How to verify: export returns for the last quarter, group by product and reason, and compare the worst products with their size chart and fit note.
Owned by: you. It is your data and your review.
8. A visual answer for the human shopper
What it is: a try-on preview on the product page, rendering the garment on the shopper's own photo.
Why it matters: even a well-answered agent query cannot show a specific shopper what a garment looks like on them. A preview helps with style and proportion. It is a visual aid and does not establish physical fit, so it sits alongside size guidance, not in place of it.
How to verify: open a live product page and confirm the try-on block appears and produces a preview.
Shared: Rendered Fits provides the block, and you place it through the Shopify theme editor.
The checklist at a glance
| Item | Who owns it | Verify with |
|---|---|---|
| 1. Structured Product/Offer schema | You | Rich Results Test |
| 2. Machine-readable size guide | You | View page source |
| 3. Fit notes as plain text | You | View page source |
| 4. Agent or API integration | Decided per store | Written confirmation from the vendor |
| 5. Fit fields in Shopify metafields | You | Shopify admin custom data |
| 6. MCP or similar interfaces | Decided per store | Specification and a test call |
| 7. Size guidance vs your returns | You | Returns by product and reason |
| 8. Visual try-on for the human | Rendered Fits + you (placement) | View a live PDP |
Frequently asked questions
How do I know if my store is AI-shopper-ready?
Work through the eight checks above. Checks 1 to 3 are catalogue hygiene you own. Checks 4 and 6 are questions to settle per store before you assume any agent or API interface exists. Checks 5, 7 and 8 cover fit fields you control, a review of size guidance against your own returns, and a visual answer for the human shopper.
Does Rendered Fits provide agent or API endpoints?
Integrations for APIs, agents or platforms are discussed per store. Rendered Fits does not present a public API contract or universal production endpoints, and earlier blanket claims about them have been removed. Contact Rendered Fits to discuss a specific store, and confirm availability, authentication, scope and data handling first.
What do I still have to do myself?
Keep core product data as Schema.org Product/Offer markup, keep your size guide and fit notes as on-page text rather than images or PDFs, review size guidance against your own returns, and place any try-on block on your product pages through the Shopify theme editor.
How long does this take?
It depends on the size and state of your catalogue. Cleaning up product schema, size charts and fit notes is ongoing catalogue hygiene rather than a one-off install, so there is no single timeframe.
Does a try-on preview prove that a garment will fit?
No. A try-on preview shows style and proportion on a shopper's photo. It does not establish physical fit. Size guidance uses your size chart, fit notes and available sizing information, and neither function guarantees fit.